INDEX

Explanations

comparisons involving multiplicative phrases or ratios

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 29-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L29R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.29.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 à¸¿

-0.15

TickCount

-0.15

YC

-0.15

essian

-0.14

lest

-0.14

ABCDEFG

-0.14

ActivityIndicatorView

-0.14

thalm

-0.14

anÃ©

-0.14

ewise

-0.13

POSITIVE LOGITS

 times

0.30

 fold

0.24

 greater

0.24

fold

0.23

 Times

0.22

 what

0.22

times

0.21

 normal

0.21

(!

0.20

 Fold

0.20

Activations Density 0.031%

comparisons involving multiplicative phrases or ratios

No Comments

No Known Activations